/*

The MIT License (MIT)

Copyright (c) 2011-2016 Broad Institute, Aiden Lab

Permission is hereby granted, free of charge, to any person obtaining a copy

of this software and associated documentation files (the "Software"), to deal

in the Software without restriction, including without limitation the rights

to use, copy, modify, merge, publish, distribute, sublicense, and/or sell

copies of the Software, and to permit persons to whom the Software is

furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in

all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR

IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,

FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE

AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER

LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,

OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN

THE SOFTWARE.

*/

#include <cstring>

#include <iostream>

#include <fstream>

#include <sstream>

#include <map>

#include <cmath>

#include <set>

#include <vector>

#include <streambuf>

#include <curl/curl.h>

#include "zlib.h"

#include "straw.h"

using namespace std;

/*

Straw: fast C++ implementation of dump. Not as fully featured as the

Java version. Reads the .hic file, finds the appropriate matrix and slice

of data, and outputs as text in sparse upper triangular format.

Currently only supporting matrices.

Usage: straw <NONE/VC/VC_SQRT/KR> <hicFile(s)> <chr1>[:x1:x2] <chr2>[:y1:y2] <BP/FRAG> <binsize>

*/

// this is for creating a stream from a byte array for ease of use

struct membuf : std::streambuf {

membuf(char *begin, char *end) {

this->setg(begin, begin, end);

}

};

// for holding data from URL call

struct MemoryStruct {

char *memory;

size_t size;

};

// callback for libcurl. data written to this buffer

static size_t

WriteMemoryCallback(void *contents, size_t size, size_t nmemb, void *userp) {

size_t realsize = size * nmemb;

struct MemoryStruct *mem = (struct MemoryStruct *) userp;

mem->memory = static_cast<char *>(realloc(mem->memory, mem->size + realsize + 1));

if (mem->memory == nullptr) {

/* out of memory! */

printf("not enough memory (realloc returned NULL)\n");

return 0;

}

std::memcpy(&(mem->memory[mem->size]), contents, realsize);

mem->size += realsize;

mem->memory[mem->size] = 0;

return realsize;

}

// get a buffer that can be used as an input stream from the URL

char *getData(CURL *curl, int64_t position, int64_t chunksize) {

std::ostringstream oss;

struct MemoryStruct chunk;

chunk.memory = static_cast<char *>(malloc(1));

chunk.size = 0; /* no data at this point */

oss << position << "-" << position + chunksize;

curl_easy_setopt(curl, CURLOPT_WRITEDATA, (void *) &chunk);

curl_easy_setopt(curl, CURLOPT_RANGE, oss.str().c_str());

CURLcode res = curl_easy_perform(curl);

if (res != CURLE_OK) {

fprintf(stderr, "curl_easy_perform() failed: %s\n",

curl_easy_strerror(res));

}

// printf("%lu bytes retrieved\n", (int64_t)chunk.size);

return chunk.memory;

}

// returns whether or not this is valid HiC file

bool readMagicString(istream &fin) {

string str;

getline(fin, str, '\0');

return str[0] == 'H' && str[1] == 'I' && str[2] == 'C';

}

char readCharFromFile(istream &fin) {

char tempChar;

fin.read(&tempChar, sizeof(char));

return tempChar;

}

int16_t readInt16FromFile(istream &fin) {

int16_t tempInt16;

fin.read((char *) &tempInt16, sizeof(int16_t));

return tempInt16;

}

int32_t readInt32FromFile(istream &fin) {

int32_t tempInt32;

fin.read((char *) &tempInt32, sizeof(int32_t));

return tempInt32;

}

int64_t readInt64FromFile(istream &fin) {

int64_t tempInt64;

fin.read((char *) &tempInt64, sizeof(int64_t));

return tempInt64;

}

float readFloatFromFile(istream &fin) {

float tempFloat;

fin.read((char *) &tempFloat, sizeof(float));

return tempFloat;

}

double readDoubleFromFile(istream &fin) {

double tempDouble;

fin.read((char *) &tempDouble, sizeof(double));

return tempDouble;

}

// reads the header, storing the positions of the normalization vectors and returning the masterIndexPosition pointer

map<string, chromosome> readHeader(istream &fin, int64_t &masterIndexPosition, string &genomeID, int32_t &numChromosomes,

int32_t &version, int64_t &nviPosition, int64_t &nviLength) {

map<string, chromosome> chromosomeMap;

if (!readMagicString(fin)) {

cerr << "Hi-C magic string is missing, does not appear to be a hic file" << endl;

masterIndexPosition = -1;

return chromosomeMap;

}

version = readInt32FromFile(fin);

if (version < 6) {

cerr << "Version " << version << " no longer supported" << endl;

masterIndexPosition = -1;

return chromosomeMap;

}

masterIndexPosition = readInt64FromFile(fin);

getline(fin, genomeID, '\0');

if (version > 8) {

nviPosition = readInt64FromFile(fin);

nviLength = readInt64FromFile(fin);

}

int32_t nattributes = readInt32FromFile(fin);

// reading and ignoring attribute-value dictionary

for (int i = 0; i < nattributes; i++) {

string key, value;

getline(fin, key, '\0');

getline(fin, value, '\0');

}

numChromosomes = readInt32FromFile(fin);

// chromosome map for finding matrixType

for (int i = 0; i < numChromosomes; i++) {

string name;

int64_t length;

getline(fin, name, '\0');

if (version > 8) {

length = readInt64FromFile(fin);

} else {

length = (int64_t) readInt32FromFile(fin);

}

chromosome chr;

chr.index = i;

chr.name = name;

chr.length = length;

chromosomeMap[name] = chr;

}

return chromosomeMap;

}

// reads the footer from the master pointer location. takes in the chromosomes,

// norm, unit (BP or FRAG) and resolution or binsize, and sets the file

// position of the matrix and the normalization vectors for those chromosomes

// at the given normalization and resolution

bool readFooter(istream &fin, int64_t master, int32_t version, int32_t c1, int32_t c2, const string &matrixType, const string &norm,

const string &unit, int32_t resolution, int64_t &myFilePos, indexEntry &c1NormEntry, indexEntry &c2NormEntry,

vector<double> &expectedValues) {

if (version > 8) {

int64_t nBytes = readInt64FromFile(fin);

} else {

int32_t nBytes = readInt32FromFile(fin);

}

stringstream ss;

ss << c1 << "_" << c2;

string key = ss.str();

int32_t nEntries = readInt32FromFile(fin);

bool found = false;

for (int i = 0; i < nEntries; i++) {

string str;

getline(fin, str, '\0');

int64_t fpos = readInt64FromFile(fin);

int32_t sizeinbytes = readInt32FromFile(fin);

if (str == key) {

myFilePos = fpos;

found = true;

}

}

if (!found) {

cerr << "File doesn't have the given chr_chr map " << key << endl;

return false;

}

if ((matrixType == "observed" && norm == "NONE") || ((matrixType == "oe" || matrixType == "expected") && norm == "NONE" && c1 != c2))

return true; // no need to read norm vector index

// read in and ignore expected value maps; don't store; reading these to

// get to norm vector index

int32_t nExpectedValues = readInt32FromFile(fin);

for (int i = 0; i < nExpectedValues; i++) {

string unit0;

getline(fin, unit0, '\0'); //unit

int32_t binSize = readInt32FromFile(fin);

int64_t nValues;

if (version > 8) {

nValues = readInt64FromFile(fin);

} else {

nValues = (int64_t) readInt32FromFile(fin);

}

bool store = c1 == c2 && (matrixType == "oe" || matrixType == "expected") && norm == "NONE" && unit0 == unit && binSize == resolution;

if (version > 8) {

for (int j = 0; j < nValues; j++) {

double v = readFloatFromFile(fin);

if (store) {

expectedValues.push_back(v);

}

}

} else {

for (int j = 0; j < nValues; j++) {

double v = readDoubleFromFile(fin);

if (store) {

expectedValues.push_back(v);

}

}

}

int32_t nNormalizationFactors = readInt32FromFile(fin);

for (int j = 0; j < nNormalizationFactors; j++) {

int32_t chrIdx = readInt32FromFile(fin);

double v;

if (version > 8) {

v = readFloatFromFile(fin);

} else {

v = readDoubleFromFile(fin);

}

if (store && chrIdx == c1) {

for (double &expectedValue : expectedValues) {

expectedValue = expectedValue / v;

}

}

}

}

if (c1 == c2 && (matrixType == "oe" || matrixType == "expected") && norm == "NONE") {

if (expectedValues.empty()) {

cerr << "File did not contain expected values vectors at " << resolution << " " << unit << endl;

return false;

}

return true;

}

nExpectedValues = readInt32FromFile(fin);

for (int i = 0; i < nExpectedValues; i++) {

string type, unit0;

getline(fin, type, '\0'); //typeString

getline(fin, unit0, '\0'); //unit

int32_t binSize = readInt32FromFile(fin);

int64_t nValues;

if (version > 8) {

nValues = readInt64FromFile(fin);

} else {

nValues = (int64_t) readInt32FromFile(fin);

}

bool store = c1 == c2 && (matrixType == "oe" || matrixType == "expected") && type == norm && unit0 == unit && binSize == resolution;

if (version > 8) {

for (int j = 0; j < nValues; j++) {

double v = readFloatFromFile(fin);

if (store) {

expectedValues.push_back(v);

}

}

} else {

for (int j = 0; j < nValues; j++) {

double v = readDoubleFromFile(fin);

if (store) {

expectedValues.push_back(v);

}

}

}

int32_t nNormalizationFactors = readInt32FromFile(fin);

for (int j = 0; j < nNormalizationFactors; j++) {

int32_t chrIdx = readInt32FromFile(fin);

double v;

if (version > 8) {

v = (double) readFloatFromFile(fin);

} else {

v = readDoubleFromFile(fin);

}

if (store && chrIdx == c1) {

for (double &expectedValue : expectedValues) {

expectedValue = expectedValue / v;

}

}

}

}

if (c1 == c2 && (matrixType == "oe" || matrixType == "expected") && norm != "NONE") {

if (expectedValues.empty()) {

cerr << "File did not contain normalized expected values vectors at " << resolution << " " << unit << endl;

return false;

}

}

// Index of normalization vectors

nEntries = readInt32FromFile(fin);

bool found1 = false;

bool found2 = false;

for (int i = 0; i < nEntries; i++) {

string normtype;

getline(fin, normtype, '\0'); //normalization type

int32_t chrIdx = readInt32FromFile(fin);

string unit1;

getline(fin, unit1, '\0'); //unit

int32_t resolution1 = readInt32FromFile(fin);

int64_t filePosition = readInt64FromFile(fin);

int64_t sizeInBytes;

if (version > 8) {

sizeInBytes = readInt64FromFile(fin);

} else {

sizeInBytes = (int64_t) readInt32FromFile(fin);

}

if (chrIdx == c1 && normtype == norm && unit1 == unit && resolution1 == resolution) {

c1NormEntry.position = filePosition;

c1NormEntry.size = sizeInBytes;

found1 = true;

}

if (chrIdx == c2 && normtype == norm && unit1 == unit && resolution1 == resolution) {

c2NormEntry.position = filePosition;

c2NormEntry.size = sizeInBytes;

found2 = true;

}

}

if (!found1 || !found2) {

cerr << "File did not contain " << norm << " normalization vectors for one or both chromosomes at "

<< resolution << " " << unit << endl;

}

return true;

}

// reads the raw binned contact matrix at specified resolution, setting the block bin count and block column count

map<int32_t, indexEntry> readMatrixZoomData(istream &fin, const string &myunit, int32_t mybinsize, float &mySumCounts,

int32_t &myBlockBinCount, int32_t &myBlockColumnCount, bool &found) {

map<int32_t, indexEntry> blockMap;

string unit;

getline(fin, unit, '\0'); // unit

readInt32FromFile(fin); // Old "zoom" index -- not used

float sumCounts = readFloatFromFile(fin); // sumCounts

readFloatFromFile(fin); // occupiedCellCount

readFloatFromFile(fin); // stdDev

readFloatFromFile(fin); // percent95

int32_t binSize = readInt32FromFile(fin);

int32_t blockBinCount = readInt32FromFile(fin);

int32_t blockColumnCount = readInt32FromFile(fin);

found = false;

if (myunit == unit && mybinsize == binSize) {

mySumCounts = sumCounts;

myBlockBinCount = blockBinCount;

myBlockColumnCount = blockColumnCount;

found = true;

}

int32_t nBlocks = readInt32FromFile(fin);

for (int b = 0; b < nBlocks; b++) {

int32_t blockNumber = readInt32FromFile(fin);

int64_t filePosition = readInt64FromFile(fin);

int32_t blockSizeInBytes = readInt32FromFile(fin);

indexEntry entry = indexEntry();

entry.size = (int64_t) blockSizeInBytes;

entry.position = filePosition;

if (found) blockMap[blockNumber] = entry;

}

return blockMap;

}

// reads the raw binned contact matrix at specified resolution, setting the block bin count and block column count

map<int32_t, indexEntry> readMatrixZoomDataHttp(CURL *curl, int64_t &myFilePosition, const string &myunit, int32_t mybinsize,

float &mySumCounts, int32_t &myBlockBinCount, int32_t &myBlockColumnCount,

bool &found) {

map<int32_t, indexEntry> blockMap;

char *buffer;

int32_t header_size = 5 * sizeof(int32_t) + 4 * sizeof(float);

char *first;

first = getData(curl, myFilePosition, 1);

if (first[0] == 'B') {

header_size += 3;

} else if (first[0] == 'F') {

header_size += 5;

} else {

cerr << "Unit not understood" << endl;

return blockMap;

}

buffer = getData(curl, myFilePosition, header_size);

membuf sbuf(buffer, buffer + header_size);

istream fin(&sbuf);

string unit;

getline(fin, unit, '\0'); // unit

readInt32FromFile(fin); // Old "zoom" index -- not used

float sumCounts = readFloatFromFile(fin); // sumCounts

readFloatFromFile(fin); // occupiedCellCount

readFloatFromFile(fin); // stdDev

readFloatFromFile(fin); // percent95

int32_t binSize = readInt32FromFile(fin);

int32_t blockBinCount = readInt32FromFile(fin);

int32_t blockColumnCount = readInt32FromFile(fin);

found = false;

if (myunit == unit && mybinsize == binSize) {

mySumCounts = sumCounts;

myBlockBinCount = blockBinCount;

myBlockColumnCount = blockColumnCount;

found = true;

}

int32_t nBlocks = readInt32FromFile(fin);

if (found) {

int32_t chunkSize = nBlocks * (sizeof(int32_t) + sizeof(int64_t) + sizeof(int32_t));

buffer = getData(curl, myFilePosition + header_size, chunkSize);

membuf sbuf2(buffer, buffer + chunkSize);

istream fin2(&sbuf2);

for (int b = 0; b < nBlocks; b++) {

int32_t blockNumber = readInt32FromFile(fin2);

int64_t filePosition = readInt64FromFile(fin2);

int32_t blockSizeInBytes = readInt32FromFile(fin2);

indexEntry entry = indexEntry();

entry.size = (int64_t) blockSizeInBytes;

entry.position = filePosition;

blockMap[blockNumber] = entry;

}

} else {

myFilePosition = myFilePosition + header_size + (nBlocks * (sizeof(int32_t) + sizeof(int64_t) + sizeof(int32_t)));

}

delete buffer;

return blockMap;

}

// goes to the specified file pointer in http and finds the raw contact matrixType at specified resolution, calling readMatrixZoomData.

// sets blockbincount and blockcolumncount

map<int32_t, indexEntry> readMatrixHttp(CURL *curl, int64_t myFilePosition, const string &unit, int32_t resolution,

float &mySumCounts, int32_t &myBlockBinCount, int32_t &myBlockColumnCount) {

char *buffer;

int32_t size = sizeof(int32_t) * 3;

buffer = getData(curl, myFilePosition, size);

membuf sbuf(buffer, buffer + size);

istream bufin(&sbuf);

int32_t c1 = readInt32FromFile(bufin);

int32_t c2 = readInt32FromFile(bufin);

int32_t nRes = readInt32FromFile(bufin);

int32_t i = 0;

bool found = false;

myFilePosition = myFilePosition + size;

delete buffer;

map<int32_t, indexEntry> blockMap;

while (i < nRes && !found) {

// myFilePosition gets updated within call

blockMap = readMatrixZoomDataHttp(curl, myFilePosition, unit, resolution, mySumCounts, myBlockBinCount, myBlockColumnCount,

found);

i++;

}

if (!found) {

cerr << "Error finding block data" << endl;

}

return blockMap;

}

// goes to the specified file pointer and finds the raw contact matrixType at specified resolution, calling readMatrixZoomData.

// sets blockbincount and blockcolumncount

map<int32_t, indexEntry> readMatrix(istream &fin, int64_t myFilePosition, const string &unit, int32_t resolution,

float &mySumCounts, int32_t &myBlockBinCount, int32_t &myBlockColumnCount) {

map<int32_t, indexEntry> blockMap;

fin.seekg(myFilePosition, ios::beg);

int32_t c1 = readInt32FromFile(fin);

int32_t c2 = readInt32FromFile(fin);

int32_t nRes = readInt32FromFile(fin);

int32_t i = 0;

bool found = false;

while (i < nRes && !found) {

blockMap = readMatrixZoomData(fin, unit, resolution, mySumCounts, myBlockBinCount, myBlockColumnCount, found);

i++;

}

if (!found) {

cerr << "Error finding block data" << endl;

}

return blockMap;

}

// gets the blocks that need to be read for this slice of the data. needs blockbincount, blockcolumncount, and whether

// or not this is intrachromosomal.

set<int32_t> getBlockNumbersForRegionFromBinPosition(const int64_t *regionIndices, int32_t blockBinCount, int32_t blockColumnCount,

bool intra) {

int32_t col1 = static_cast<int32_t>(regionIndices[0] / blockBinCount);

int32_t col2 = static_cast<int32_t>((regionIndices[1] + 1) / blockBinCount);

int32_t row1 = static_cast<int32_t>(regionIndices[2] / blockBinCount);

int32_t row2 = static_cast<int32_t>((regionIndices[3] + 1) / blockBinCount);

set<int32_t> blocksSet;

// first check the upper triangular matrixType

for (int r = row1; r <= row2; r++) {

for (int c = col1; c <= col2; c++) {

int32_t blockNumber = r * blockColumnCount + c;

blocksSet.insert(blockNumber);

}

}

// check region part that overlaps with lower left triangle but only if intrachromosomal

if (intra) {

for (int r = col1; r <= col2; r++) {

for (int c = row1; c <= row2; c++) {

int32_t blockNumber = r * blockColumnCount + c;

blocksSet.insert(blockNumber);

}

}

}

return blocksSet;

}

set<int32_t> getBlockNumbersForRegionFromBinPositionV9Intra(int64_t *regionIndices, int32_t blockBinCount, int32_t blockColumnCount) {

// regionIndices is binX1 binX2 binY1 binY2

set<int32_t> blocksSet;

int32_t translatedLowerPAD = static_cast<int32_t>((regionIndices[0] + regionIndices[2]) / 2 / blockBinCount);

int32_t translatedHigherPAD = static_cast<int32_t>((regionIndices[1] + regionIndices[3]) / 2 / blockBinCount + 1);

int32_t translatedNearerDepth = static_cast<int32_t>(log2(

1 + abs(regionIndices[0] - regionIndices[3]) / sqrt(2) / blockBinCount));

int32_t translatedFurtherDepth = static_cast<int32_t>(log2(

1 + abs(regionIndices[1] - regionIndices[2]) / sqrt(2) / blockBinCount));

// because code above assume above diagonal; but we could be below diagonal

int32_t nearerDepth = min(translatedNearerDepth, translatedFurtherDepth);

if ((regionIndices[0] > regionIndices[3] && regionIndices[1] < regionIndices[2]) ||

(regionIndices[1] > regionIndices[2] && regionIndices[0] < regionIndices[3])) {

nearerDepth = 0;

}

int32_t furtherDepth = max(translatedNearerDepth, translatedFurtherDepth) + 1; // +1; integer divide rounds down

for (int depth = nearerDepth; depth <= furtherDepth; depth++) {

for (int pad = translatedLowerPAD; pad <= translatedHigherPAD; pad++) {

int32_t blockNumber = depth * blockColumnCount + pad;

blocksSet.insert(blockNumber);

}

}

return blocksSet;

}

void appendRecord(vector<contactRecord> &vector, int32_t index, int32_t binX, int32_t binY, float counts) {

contactRecord record = contactRecord();

record.binX = binX;

record.binY = binY;

record.counts = counts;

vector[index] = record;

}

// this is the meat of reading the data. takes in the block number and returns the set of contact records corresponding to

// that block. the block data is compressed and must be decompressed using the zlib library functions

vector<contactRecord> readBlock(istream &fin, CURL *curl, bool isHttp, indexEntry idx, int32_t version) {

if (idx.size <= 0) {

vector<contactRecord> v;

return v;

}

char *compressedBytes = new char[idx.size];

char *uncompressedBytes = new char[idx.size * 10]; //biggest seen so far is 3

if (isHttp) {

compressedBytes = getData(curl, idx.position, idx.size);

} else {

fin.seekg(idx.position, ios::beg);

fin.read(compressedBytes, idx.size);

}

// Decompress the block

// zlib struct

z_stream infstream;

infstream.zalloc = Z_NULL;

infstream.zfree = Z_NULL;

infstream.opaque = Z_NULL;

infstream.avail_in = static_cast<uInt>(idx.size); // size of input

infstream.next_in = (Bytef *) compressedBytes; // input char array

infstream.avail_out = static_cast<uInt>(idx.size * 10); // size of output

infstream.next_out = (Bytef *) uncompressedBytes; // output char array

// the actual decompression work.

inflateInit(&infstream);

inflate(&infstream, Z_NO_FLUSH);

inflateEnd(&infstream);

int32_t uncompressedSize = static_cast<int32_t>(infstream.total_out);

// create stream from buffer for ease of use

membuf sbuf(uncompressedBytes, uncompressedBytes + uncompressedSize);

istream bufferin(&sbuf);

uint64_t nRecords = static_cast<uint64_t>(readInt32FromFile(bufferin));

vector<contactRecord> v(nRecords);

// different versions have different specific formats

if (version < 7) {

for (int i = 0; i < nRecords; i++) {

int32_t binX = readInt32FromFile(bufferin);

int32_t binY = readInt32FromFile(bufferin);

float counts = readFloatFromFile(bufferin);

appendRecord(v, i, binX, binY, counts);

}

} else {

int32_t binXOffset = readInt32FromFile(bufferin);

int32_t binYOffset = readInt32FromFile(bufferin);

bool useShort = readCharFromFile(bufferin) == 0; // yes this is opposite of usual

bool useShortBinX = true;

bool useShortBinY = true;

if (version > 8) {

useShortBinX = readCharFromFile(bufferin) == 0;

useShortBinY = readCharFromFile(bufferin) == 0;

}

char type = readCharFromFile(bufferin);

int32_t index = 0;

if (type == 1) {

if (useShortBinX && useShortBinY) {

int16_t rowCount = readInt16FromFile(bufferin);

for (int i = 0; i < rowCount; i++) {

int32_t binY = binYOffset + readInt16FromFile(bufferin);

int16_t colCount = readInt16FromFile(bufferin);

for (int j = 0; j < colCount; j++) {

int32_t binX = binXOffset + readInt16FromFile(bufferin);

float counts;

if (useShort) {

counts = readInt16FromFile(bufferin);

} else {

counts = readFloatFromFile(bufferin);

}

appendRecord(v, index++, binX, binY, counts);

}

}

} else if (useShortBinX && !useShortBinY) {

int32_t rowCount = readInt32FromFile(bufferin);

for (int i = 0; i < rowCount; i++) {

int32_t binY = binYOffset + readInt32FromFile(bufferin);

int16_t colCount = readInt16FromFile(bufferin);

for (int j = 0; j < colCount; j++) {

int32_t binX = binXOffset + readInt16FromFile(bufferin);

float counts;

if (useShort) {

counts = readInt16FromFile(bufferin);

} else {

counts = readFloatFromFile(bufferin);

}

appendRecord(v, index++, binX, binY, counts);

}

}

} else if (!useShortBinX && useShortBinY) {

int16_t rowCount = readInt16FromFile(bufferin);

for (int i = 0; i < rowCount; i++) {

int32_t binY = binYOffset + readInt16FromFile(bufferin);

int32_t colCount = readInt32FromFile(bufferin);

for (int j = 0; j < colCount; j++) {

int32_t binX = binXOffset + readInt32FromFile(bufferin);

float counts;

if (useShort) {

counts = readInt16FromFile(bufferin);

} else {

counts = readFloatFromFile(bufferin);

}

appendRecord(v, index++, binX, binY, counts);

}

}

} else {

int32_t rowCount = readInt32FromFile(bufferin);

for (int i = 0; i < rowCount; i++) {

int32_t binY = binYOffset + readInt32FromFile(bufferin);

int32_t colCount = readInt32FromFile(bufferin);

for (int j = 0; j < colCount; j++) {

int32_t binX = binXOffset + readInt32FromFile(bufferin);

float counts;

if (useShort) {

counts = readInt16FromFile(bufferin);

} else {

counts = readFloatFromFile(bufferin);

}

appendRecord(v, index++, binX, binY, counts);

}

}

}

} else if (type == 2) {

int32_t nPts = readInt32FromFile(bufferin);

int16_t w = readInt16FromFile(bufferin);

for (int i = 0; i < nPts; i++) {

//int32_t idx = (p.y - binOffset2) * w + (p.x - binOffset1);

int32_t row = i / w;

int32_t col = i - row * w;

int32_t bin1 = binXOffset + col;

int32_t bin2 = binYOffset + row;

float counts;

if (useShort) {

int16_t c = readInt16FromFile(bufferin);

if (c != -32768) {

appendRecord(v, index++, bin1, bin2, c);

}

} else {

counts = readFloatFromFile(bufferin);

if (!isnan(counts)) {

appendRecord(v, index++, bin1, bin2, counts);

}

}

}

}

}

delete[] compressedBytes;

delete[] uncompressedBytes; // don't forget to delete your heap arrays in C++!

return v;

}

// reads the normalization vector from the file at the specified location

vector<double> readNormalizationVector(istream &bufferin, int32_t version) {

int64_t nValues;

if (version > 8) {

nValues = readInt64FromFile(bufferin);

} else {

nValues = (int64_t) readInt32FromFile(bufferin);

}

uint64_t numValues = static_cast<uint64_t>(nValues);

vector<double> values(numValues);

if (version > 8) {

for (int i = 0; i < nValues; i++) {

values[i] = (double) readFloatFromFile(bufferin);

}

} else {

for (int i = 0; i < nValues; i++) {

values[i] = readDoubleFromFile(bufferin);

}

}

return values;

}

class FileReader {

public:

string prefix = "http"; // HTTP code

ifstream fin;

CURL *curl;

bool isHttp = false;

static CURL *initCURL(const char *url) {

CURL *curl = curl_easy_init();

if (curl) {

curl_easy_setopt(curl, CURLOPT_WRITEFUNCTION, WriteMemoryCallback);

curl_easy_setopt(curl, CURLOPT_URL, url);

curl_easy_setopt(curl, CURLOPT_FOLLOWLOCATION, 1L);

curl_easy_setopt(curl, CURLOPT_USERAGENT, "straw");

}

return curl;

}

explicit FileReader(const string &fname) {

// read header into buffer; 100K should be sufficient

if (std::strncmp(fname.c_str(), prefix.c_str(), prefix.size()) == 0) {

isHttp = true;

curl = initCURL(fname.c_str());

if (!curl) {

cerr << "URL " << fname << " cannot be opened for reading" << endl;

exit(1);

}

} else {

fin.open(fname, fstream::in | fstream::binary);

if (!fin) {

cerr << "File " << fname << " cannot be opened for reading" << endl;

exit(2);

}

}

}

void close(){

if(isHttp){

curl_easy_cleanup(curl);

} else {

fin.close();

}

}

};

class HiCFile {

public:

string prefix = "http"; // HTTP code

bool isHttp = false;

ifstream fin;

CURL *curl;

int64_t master = 0LL;

map<string, chromosome> chromosomeMap;

string genomeID;

int32_t numChromosomes = 0;

int32_t version = 0;

int64_t nviPosition = 0LL;

int64_t nviLength = 0LL;

static int64_t totalFileSize;

static size_t hdf(char *b, size_t size, size_t nitems, void *userdata) {

size_t numbytes = size * nitems;

b[numbytes + 1] = '\0';

string s(b);

int32_t found = static_cast<int32_t>(s.find("Content-Range"));

if (found != string::npos) {

int32_t found2 = static_cast<int32_t>(s.find("/"));

//Content-Range: bytes 0-100000/891471462

if (found2 != string::npos) {

string total = s.substr(found2 + 1);

totalFileSize = stol(total);

}

}

return numbytes;

}

static CURL *initCURL(const char *url) {

CURL *curl = curl_easy_init();

if (curl) {

curl_easy_setopt(curl, CURLOPT_WRITEFUNCTION, WriteMemoryCallback);

curl_easy_setopt(curl, CURLOPT_URL, url);

curl_easy_setopt(curl, CURLOPT_FOLLOWLOCATION, 1L);

curl_easy_setopt(curl, CURLOPT_HEADERFUNCTION, hdf);

curl_easy_setopt(curl, CURLOPT_USERAGENT, "straw");

}

return curl;

}

explicit HiCFile(const string &fname) {

// read header into buffer; 100K should be sufficient

if (std::strncmp(fname.c_str(), prefix.c_str(), prefix.size()) == 0) {

isHttp = true;

char *buffer;

curl = initCURL(fname.c_str());

if (curl) {

buffer = getData(curl, 0, 100000);

} else {

cerr << "URL " << fname << " cannot be opened for reading" << endl;

exit(1);

}

membuf sbuf(buffer, buffer + 100000);

istream bufin(&sbuf);

chromosomeMap = readHeader(bufin, master, genomeID, numChromosomes,

version, nviPosition, nviLength);

delete buffer;

} else {

fin.open(fname, fstream::in | fstream::binary);

if (!fin) {

cerr << "File " << fname << " cannot be opened for reading" << endl;

exit(2);

}

chromosomeMap = readHeader(fin, master, genomeID, numChromosomes,

version, nviPosition, nviLength);

}

}

void close(){

if(isHttp){

curl_easy_cleanup(curl);

} else {

fin.close();

}

}

vector<double> readNormalizationVectorFromFooter(indexEntry cNormEntry) {

char *buffer;

if (isHttp) {

buffer = getData(curl, cNormEntry.position, cNormEntry.size);

} else {

buffer = new char[cNormEntry.size];

fin.seekg(cNormEntry.position, ios::beg);

fin.read(buffer, cNormEntry.size);

}

membuf sbuf3(buffer, buffer + cNormEntry.size);

istream bufferin(&sbuf3);

vector<double> cNorm = readNormalizationVector(bufferin, version);

delete buffer;

return cNorm;

}

};

int64_t HiCFile::totalFileSize = 0LL;

class MatrixZoomData {

public:

indexEntry c1NormEntry, c2NormEntry;

int64_t myFilePos = 0LL;

vector<double> expectedValues;

bool foundFooter = false;

vector<double> c1Norm;

vector<double> c2Norm;

int32_t c1 = 0;

int32_t c2 = 0;

string matrixType;

string norm;

string unit;

int32_t resolution = 0;

int32_t numBins1 = 0;

int32_t numBins2 = 0;

MatrixZoomData(HiCFile *hiCFile, const chromosome &chrom1, const chromosome &chrom2, const string &matrixType,

const string &norm, const string &unit, int32_t resolution) {

int32_t c01 = chrom1.index;

int32_t c02 = chrom2.index;

if (c01 <= c02) { // default is ok

this->c1 = c01;

this->c2 = c02;

this->numBins1 = static_cast<int32_t>(chrom1.length / resolution);

this->numBins2 = static_cast<int32_t>(chrom2.length / resolution);

} else { // flip

this->c1 = c02;

this->c2 = c01;

this->numBins1 = static_cast<int32_t>(chrom2.length / resolution);

this->numBins2 = static_cast<int32_t>(chrom1.length / resolution);

}

this->matrixType = matrixType;

this->norm = norm;

this->unit = unit;

this->resolution = resolution;

if (hiCFile->isHttp) {

char *buffer2;

int64_t bytes_to_read = hiCFile->totalFileSize - hiCFile->master;

buffer2 = getData(hiCFile->curl, hiCFile->master, bytes_to_read);

membuf sbuf2(buffer2, buffer2 + bytes_to_read);

istream bufin2(&sbuf2);

foundFooter = readFooter(bufin2, hiCFile->master, hiCFile->version, c1, c2, matrixType, norm, unit,

resolution,

myFilePos,

c1NormEntry, c2NormEntry, expectedValues);

delete buffer2;

} else {

hiCFile->fin.seekg(hiCFile->master, ios::beg);

foundFooter = readFooter(hiCFile->fin, hiCFile->master, hiCFile->version, c1, c2, matrixType, norm,

unit,

resolution, myFilePos,

c1NormEntry, c2NormEntry, expectedValues);

}

if (!foundFooter) {

return;

}

if (norm != "NONE") {

c1Norm = hiCFile->readNormalizationVectorFromFooter(c1NormEntry);

if (c1 == c2) {

c2Norm = c1Norm;

} else {

c2Norm = hiCFile->readNormalizationVectorFromFooter(c2NormEntry);

}